Robotic colorectal surgery: more than a fantastic toy?

There has been a rapid rise in the number of robotic colorectal procedures worldwide since the da Vinci Surgical System robotic technology was approved for surgical procedures in the year 2000. Several recent meta-analyses and systematic reviews have shown a significant difference in outcomes between robotic and laparoscopic rectal cancer surgery. However, these results from pooled data have not been supported by the initial results reported from the Robotic assisted versus laparoscopic assisted resection for rectal cancer trial. In this article, we examine the current evidence for robotic colorectal surgery, assess its features and functionality, evaluate its learning curve and provide our perspective on its future. ©2018 Toh J.W.T. et al., published by De Gruyter, Berlin/Boston.

There has been a rapid rise in the number of robotic colorectal procedures worldwide since the da Vinci Surgical System (Intuitive Surgical, Sunnyvale, CA, USA) robotic technology was approved for surgical procedures in the year 2000. However, robust clinical evidence supporting the benefit of robotic technology in the context of colorectal surgery remains limited.

A meta-analysis comparing laparoscopic and robotic colorectal surgery by Trastulli et al. [1] reported no significant difference in operation time, length of hospital stay, morbidity and mortality, nor improvements in the quality of resection specimens. The only reported benefit reported by Trastulli et al. [1] was a reduction in conversion rate with robotic surgery (2% vs. 7.5%, p=0.0007).

In contrast, several recent meta-analyses and systematic reviews have shown a significant difference in outcomes between robotic and laparoscopic rectal cancer surgery (Table 1). Sun et al. [2], Wang et al. [3] and Xiong et al. [4] have all reported lower rates of positive circumferential resection margin involvement, erectile dysfunction and conversion rate with robotic rectal cancer surgery. The study by Sun et al. [2] also reported shorter hospital stays and lower overall complication rates.

Table 1:

Key level evidence comparing outcomes of robotic vs. laparoscopic surgery for rectal cancer.

First author	Year	Journal	Study design	Level of evidence	CRM involvement	TME quality	Conversion to open
Jayne	2017	JAMA	Randomized controlled trial (n=471; 237 robotic, 234 laparoscopic)	Level 2	Similar (OR=0.78, p=0.56)	–	Similar (OR=0.61, p=0.16)
Prete	2017	Ann Surg	Meta-analysis (n=681 from 5 studies)	Level 1a	Similar	–	Lower in robotic: 7.3% (RR=0.58, 95% CI 0.35–0.97, p=0.04)
Li	2017	Medicine (Baltimore)	Meta-analysis (n=3601 from 17 studies)	Level 1b	Similar (OR=0.80, p=0.256)	–	Lower in robotic: 7.3% (OR=0.35, p<0.001)
Sun	2016	World J Surg Oncol	Meta-analysis (n=592 from 8 studies)	Level 1b	Lower in robotic (OR=0.5, 95% CI 0.25–1.01, p=0.05)	–	Lower in robotic (OR=0.08, 95% CI 0.02–0.31, p=0.0002)
Wang	2016	Surg Laparosc Endosc Percutan Tech	Meta-analysis (n=1229 from 8 studies)	Level 1b	Lower in robotic (OR=0.44, 95% CI 0.20–0.96, p<0.05)	–	Lower in robotic (OR=0.23, 95% CI 0.10–0.52, p<0.01)
Speicher	2015	Ann Surg	US National Cancer Database (n=6403 in 2011–2012; 956 robotic)	Level 3b	Similar (5.5% vs. 4.7%)	–	Lower in robotic (9.5% vs. 16.4%, p<0.001)
Xiong	2015	J Gastrointest Surg	Meta-analysis (n=1229 from 8 studies)	Level 1b	Lower in robotic (2.7% vs. 5.8%; OR=0.44, 95% CI 0.20–0.96, p=0.04)	–	Lower in robotic (OR=0.23, 95% CI 0.10–0.52, p=0.0004)

CRM, circumferential resection margin; TME, total mesorectal excision; OR, odds ratio; CI, confidence interval; RR, relative risk; –, not reported.

These results from pooled data have not, however, been supported by the initial results reported from the ROLARR (robotic-assisted versus laparoscopic-assisted resection for rectal cancer) trial [5, 6]. This trial has not demonstrated any statistically significant differences between laparoscopic- and robotic-assisted surgery [5, 6], with its long-term results widely anticipated.

A number of studies have also reported long-term 5-year survival data for robotic colorectal procedures in comparison with laparoscopic and open approaches (Table 2). The literature has been predominantly limited to level 3 evidence thus far. Matched comparisons by Kim et al. [7] and Cho et al. [8] have not shown statistically significant differences in 5-year overall survival or disease-free survival between robotic and laparoscopic colorectal surgery, although Kim et al. [7] did demonstrate a trend to improved survival.

Table 2:

Key long-term 5-year survival data for robotic vs. laparoscopic surgery for rectal cancer.

First author	Year	Journal	Study design	Level of evidence	Local recurrence	5-Year OS	5-Year DFS	5-Year CSS
Sujatha-Bhaskar	2017	Ann Surg	US National Cancer Database (905 robotic vs. 2009 lap vs. 3399 open)	Level 3	–	78% (robotic) vs. 81% (lap) vs. 76% (open), p=0.0198	–	–
Kim	2017	DCR	Single centre (224 robotic vs. 224 lap)	Level 3	5.6% vs. 7.3% (p=0.502)	91% vs. 78% (p=NS)	73% vs. 68% (p=NS)	91% vs. 80% (p=NS)
Sammour	2017	Ann Surg	Single centre (276 robotic)	Level 3	2.40%	87%	82%	–
Cho	2015	Medicine (Baltimore)	Single centre (278 robotic vs. 278 lap)	Level 3	5.9% vs. 3.9% (p=0.313)	92% vs. 93% (p=NS)	82% vs. 80% (p=NS)	–
Ghezzi	2014	EJSO	Two centres (65 robotic vs. 109 open)	Level 3	3.4% vs. 16.1% (p=0.024)	85% vs. 76% (p=NS)	73% vs. 70% (p=NS)	–

OS, overall survival; DFS, disease-free survival; CSS, cancer-specific survival; lap, laparoscopic; NS, not significant.

It appears that the theoretical technical advantages in robotic surgery, such as with instruments that can rotate and bend in all directions, three-dimensional high-definition vision and surgeon-controlled multi-arms have not translated to huge benefits over traditional laparoscopic surgery. Opponents of robotic surgery believe that it is just a fantastic toy. We believe not.

First, the lack of reported benefits of robotic surgery over laparoscopic surgery may be associated with the general reporting of technically straightforward cases alongside technically difficult rectal cases. This may skew the results towards no difference, resulting in a type II error for technically difficult rectal cases. It is also important to understand the generalisability of pooled data – and to analyse data for specific patient subgroups, particularly for robotic surgery in order to justify its expense, as the major advantage of robotic technology is confined to a smaller cohort of patients with low rectal tumours, narrow pelvis and obesity [5]. The consensus statement regarding the use of robotics in general surgery has surmised that the advantage of robotic surgery is mainly within confined spaces [9].

Patient selection is the key. With the use of the procedure-targeted database of the American College of Surgeons National Surgical Quality Improvement Program, after matching patients based on gender, body mass index, surgical procedure, diagnosis and American Society of Anesthesiologists classification, Benlice et al. [10] reported shorter hospital stay, lower morbidity, superficial surgical site infection, transfusion requirement and ileus associated with robotic surgery over traditional laparoscopic surgery in matched patients.

In technically challenging cases, most surgeons who perform both robotic and laparoscopic surgery would anecdotally agree that robotic technology protects against conversion to open, with the superiority of robotic instruments over laparoscopic rigid instruments and significantly better vision of the surgical field associated with robotic technology. The better stereo-optic three-dimensional vision, self-controlling camera and instruments with seven degrees of freedom and motion scaling, as well as ergonomic comfort for the surgeon improve the accuracy of surgical dissection. A recent study by Ahmed et al. [11] focusing on high-risk patients (body mass index ≥30, male gender, preoperative chemoradiotherapy, tumour <8 cm from the anal verge, previous abdominal surgery) demonstrated that robotic surgery in high-risk patients resulted in higher rates of sphincter preservation; lower conversion rates, operating times and blood loss; and shorter length of stay.

However, it would be a mistake to reserve robotic technology only for complex cases, as there is a steep learning curve and longer operating time associated with learning to use robotic technology. The number of cases required for expertise in robotic surgery has been reported in several studies. There is a wide discrepancy within the literature. Herrell and Smith [12] indicated that at least 150 cases were required for proficiency. Meanwhile, the study by Ahlering et al. [13] suggested that only 9–12 cases were required to achieve proficiency. Ou et al. [14] reported that the incidence of positive surgical margins in robotic surgery decreased significantly after 250 cases. There needs to be sufficient caseload for training, credentialing and maintaining robotic skills. Some argue that the theoretical advantage of robotic surgery over laparoscopic surgery is skewed towards no difference due to the learning curve associated with robotic surgery, as most colorectal surgeons in this era are experienced laparoscopic surgeons and novice robotic surgeons. With time and experience, the operation time and other parameters associated with robotic surgery should improve, and the advantage of robotic surgery may be realised.

At our institution, we reserve robotic technology mainly for middle and low rectal cancers, and we do not use robotic technology for purely colonic nor for upper rectal cancer cases. A randomised clinical trial by Park et al. [15] showed that robotic-assisted laparoscopic right hemicolectomy provided no significant benefit over conventional laparoscopic right hemicolectomy to justify its cost. We do not believe that there is any major theoretical nor reported clinical benefit for robotics in purely colonic cases, and use robotics in selected upper rectal cases only.

Apart from improvements in short-term outcomes and conversion rates, using our own data and experience, we reported on the long-term outcomes of robotic and laparoscopic total mesorectal excision for mid and low rectal cancers (robotic, n=272; laparoscopic, n=460), showing that the robotic procedure is a positive prognostic factor in rectal cancer surgery [7]. Our study was the first in the literature to do so. After propensity scoring, we showed that our technique of single-docking totally robotic surgery was associated with a trend to improved survival. Also, in multivariate analysis, robotic surgery was a significant prognostic factor for overall survival and cancer-specific survival.

Multiple platforms are already capable of utilizing robotic surgery. These include single-incision/port surgery, robotic natural orifice transluminal endoscopic surgery, robotic transanal surgery and endo-robotics. Alongside the well-known da Vinci platform, new commercial biomedical companies including Titan Medical Inc., Endomina, Medrobotics, Flex-robotics, Memic Innovative Surgery, TransEnterix, Robot Surgeon and Auris Robotics are emerging, and we look forward to next-generation robotic systems, which represent an exciting frontier.

A toy is defined as a thing of little or no value or importance, something that does not have serious practical use, but rather is a trifle that diverts attention, mainly for amusement but with no real practical importance. We believe that robotic colorectal surgery is far from that. Rather, if we were to use a sporting metaphor, the robot is more like an impact player reserved for important rounds. The impact player makes the biggest difference when it counts, rather than in easy victories. However, it would be a strategic mistake to bring out the impact player only in the grand finale, as the team needs time and experience to work with the impact player to achieve and ensure the best outcomes.

Click here for additional data file.